Analysis of Thai food topics from Youtube transcripts using LDA and NMF techniques
Article Sidebar
Main Article Content
Abstract
This study aims to analyze the views of international users on Thai food through the transcripts in YouTube videos by comparing the performance of Latent Dirichlet Allocation (LDA) and Non-negative Matrix Factorization (NMF) techniques in identifying Thai food-related topics. Data was collected from 352 YouTube videos published between 2010-2024, data preprocessed, and transformed into N-Grams before analysis. The results revealed that LDA achieved optimal performance with the highest coherence score of 0.794 at 5 topics: (1) ingredients and food flavors, (2) tourism and food experiences, (3) Thai food components, (4) street food and popularity, and (5) cooking processes and ingredients. Meanwhile, NMF yielded optimal results with the highest coherence score of 0.956 at 8 topics: (1) seafood ingredients and flavors, (2) taste preferences, (3) street food culture, (4) Thai curry components, (5) dining experiences and food flavors, (6) food exploration, (7) rice and meat dishes, and (8) restaurant experiences and service. From the overall results of both techniques, international users perceive Thai food in four main dimensions: distinctive flavors, specific ingredients, dining experiences, and environmental context. Comparing both techniques, LDA excelled in showing connections between Thai food components through keyword overlap but had limitations in distinguishing subtopics. Meanwhile, NMF demonstrated superiority in identifying specific topics but lacked connections between them. The choice of technique should depend on the intended application, and combining results from both techniques may provide the most comprehensive perspective for developing strategies to promote Thai food internationally.
Article Details
References
Prapasawasdi U, Wuttisittikulkij L, Borompichaichartkul C, Changkaew L, Saadi M. Cultural tourism behaviors: enhancing the influence of tourists’ perceptions on local Thai Food and Culture. The open psychology journals. 2018 Oct 30;11(1).
สำนักงานนโยบายและยุทธศาสตร์การค้า, กระทรวงพาณิชย์. แนวทางการขยายตลาดสินค้าอาหารไทย. กรุงเทพฯ: สำนักงาน; 2567.
Zhang S, Lopez D, Lam L, Yenumula A, Vu T. A comparative analysis of text mining methodologies for online consumer reviews. Journal of Applied Business and Economics. 2023;25(7): doi:10.33423/jabe.v25i7.6724.
เณรัญชรา กิจวิกรานต์. ภาพลักษณ์อาหารไทย การรับรู้คุณภาพอาหารไทยและแนวโน้มพฤติกรรม การท่องเที่ยวเชิงอาหารของนักท่องเที่ยวชาวต่างประเทศ. วารสารวิชาการการท่องเที่ยวไทยนานาชาติ. 2558;10(1):12–28.
ธัมมามนตร์ คุณรัตนากรณ์, สมจิตร ล้วนจำเริญ, ราณี อิสิชัยกุล, วัชรีภรณ์ ไชยมงคล. อิทธิพลของส่วนประสมการตลาด และภาพลักษณ์ของผลิตภัณฑ์การท่องเที่ยวเชิงวัฒนธรรมที่มีต่อความภักดีในการท่องเที่ยวเพื่อเรียนรู้วัฒนธรรมไทยของนักท่องเที่ยวชาวต่างประเทศ. วารสารดุษฎีบัณฑิตทางสังคมศาสตร์. 2557;4(3):100-117.
Blei DM, Ng AY, Jordan MI. Latent dirichlet allocation. Journal of Machine Learning Research. 2003; 3:993-1022.
Seung D, Lee L. Algorithms for non-negative matrix factorization. Advances in Neural Information Processing Systems. 2001;13: 3.
Aggarwal CC, Zhai C. An Introduction to Text Mining. In Mining Text Data 2012 Jan 7 (pp. 1-10). Boston, MA: Springer US.
Yadav A, Vishwakarma DK. Sentiment analysis using deep learning architectures: a review. Artificial Intelligence Review. 2020;53(6):4335-4385.
Feldman R, Sanger J. The Text Mining Handbook: Advanced Approaches in Analyzing Unstructured Data. Cambridge university press; 2007.
Cifci I, Atsız O, Gupta V. The street food experiences of the local-guided tour in the meal-sharing economy: the case of Bangkok. British Food Journal. 2021; 123(12):4030-4048.
Qader WA, Ameen MM, Ahmed BI. An overview of bag of words; importance, implementation, applications, and challenges. In 2019 international engineering conference (IEC). 2019 Jun 23 (pp. 200-204). IEEE.
Scikit-learn: Machine Learning in Python [Internet]. Version 1.3.2. [cited 2024 Dec 11]. Available from: https://scikit-learn.org/stable/ modules/decomposition.html
Chen Y, Zhang H, Liu R, Ye Z, Lin J. Experimental explorations on short text topic mining between LDA and NMF based Schemes. Knowledge-Based Systems. 2019 Jan 1;16.
NLTK: Natural Language Toolkit [Internet]. [cited 2024 Dec 11]. Available from: https://www. nltk.org/.
Gensim: Topic Modeling for Human. [cited 2024 Dec 11]. Available from: https://radimrehurek. com/genism.
George S, Srividhya V. Implicit aspect based sentiment analysis for restaurant review using LDA topic modeling and ensemble approach. International Journal of Advanced Technology and Engineering Exploration. 2023;10(102): 554 –568.
Promsivapallop P, Kannaovakun P. Destination food image dimensions and their effects on food preference and consumption. Journal of destination marketing & management. 2019; 11: 89-100.
Chavarria LC, Phakdee-Auksorn P. Understanding international tourists' attitudes towards street food in Phuket, Thailand. Tourism Management Perspectives. 2017; 21:66-73.
Antón C, Camarero C, Laguna M, Buhalis D. Impacts of authenticity, degree of adaptation and cultural contrast on travellers’ memorable gastronomy experiences. Journal of Hospitality Marketing & Management. 2019;28(7):743–764.
